Date: Wed, 1 Jan 2003 14:26:17 -0800 (PST) From: Marcel Moolenaar <marcel@FreeBSD.org> To: Perforce Change Reviews <perforce@freebsd.org> Subject: PERFORCE change 23031 for review Message-ID: <200301012226.h01MQHEa035401@repoman.freebsd.org>
next in thread | raw e-mail | index | archive | help
http://perforce.freebsd.org/chv.cgi?CH=23031 Change 23031 by marcel@marcel_nfs on 2003/01/01 14:25:52 I cannot convince myself that reloading cr.itm with a delta of cr.itm and not a delta of cr.itc without the logic to detect missed interrupts is not going to harm. Revert the change and add a comment. We should really use cr.itm, but we need to check for missed interrupts so that we don't reload with a value that is smaller than the current counter and thus don't get timer interrupts until the 64-bit counter wraps. Affected files ... .. //depot/projects/ia64/sys/ia64/ia64/interrupt.c#22 edit Differences ... ==== //depot/projects/ia64/sys/ia64/ia64/interrupt.c#22 (text+ko) ==== @@ -128,14 +128,16 @@ /* Only the BSP runs the real clock */ if (PCPU_GET(cpuid) == 0) { #endif - ia64_set_itm(ia64_get_itm() + itm_reload); + /* XXX don't reload ITM based on ITC. */ + ia64_set_itm(ia64_get_itc() + itm_reload); hardclock((struct clockframe *)framep); /* divide hz (1024) by 8 to get stathz (128) */ if ((++schedclk2 & 0x7) == 0) statclock((struct clockframe *)framep); #ifdef SMP } else { - ia64_set_itm(ia64_get_itm() + itm_reload); + /* XXX don't reload ITM based on ITC. */ + ia64_set_itm(ia64_get_itc() + itm_reload); mtx_lock_spin(&sched_lock); hardclock_process(curthread, TRAPF_USERMODE(framep)); if ((schedclk2 & 0x7) == 0) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe p4-projects" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200301012226.h01MQHEa035401>